U.S. patent application number 12/617510 was filed with the patent office on 2010-05-13 for intensity-modulated, cone-beam computed tomographic imaging system, methods, and apparatus.
This patent application is currently assigned to THE METHODIST HOSPITAL RESEARCH INSTITUTE. Invention is credited to Goetz Benndorf, King C. Li, Xiaobo Zhou.
Application Number | 20100119033 12/617510 |
Document ID | / |
Family ID | 42165215 |
Filed Date | 2010-05-13 |
United States Patent
Application |
20100119033 |
Kind Code |
A1 |
Li; King C. ; et
al. |
May 13, 2010 |
INTENSITY-MODULATED, CONE-BEAM COMPUTED TOMOGRAPHIC IMAGING SYSTEM,
METHODS, AND APPARATUS
Abstract
Disclosed are methods for reconstructing a three-dimensional
image of an object's volume of interest using computed tomography
that employs conical-beam, intensity-modulated projections of this
object. In one embodiment, a plurality of collimating devices
serves to modulate the aperture of the radiation source thereby
acting to modulate the intensity of the source upon the object.
Also provided are image processing devices, examination apparatus,
as well as a computer-readable medium and a program element adapted
and configured to perform aspects of the methods disclosed
herein.
Inventors: |
Li; King C.; (Houston,
TX) ; Zhou; Xiaobo; (Bellaire, TX) ; Benndorf;
Goetz; (Houston, TX) |
Correspondence
Address: |
HAYNES AND BOONE, LLP;IP Section
2323 Victory Avenue, Suite 700
Dallas
TX
75219
US
|
Assignee: |
THE METHODIST HOSPITAL RESEARCH
INSTITUTE
Houston
TX
|
Family ID: |
42165215 |
Appl. No.: |
12/617510 |
Filed: |
November 12, 2009 |
Related U.S. Patent Documents
|
|
|
|
|
|
Application
Number |
Filing Date |
Patent Number |
|
|
61113655 |
Nov 12, 2008 |
|
|
|
Current U.S.
Class: |
378/5 ; 378/147;
378/4 |
Current CPC
Class: |
A61B 6/027 20130101;
A61B 6/06 20130101 |
Class at
Publication: |
378/5 ; 378/147;
378/4 |
International
Class: |
H05G 1/60 20060101
H05G001/60; G21K 1/02 20060101 G21K001/02 |
Claims
1. A method for operating computed tomographic imaging using a
radiation source and a plurality of detectors to generate an image
of an object, the method comprising the steps of: a) defining
desired image characteristics; b) performing calculations to
determine the modulation intensity to be applied to the radiation
source by at least a first aperture modulator or collimator to
generate the desired image characteristics; and c) modulating the
radiation source using the at least a first collimator to generate
a desired pattern of fluence between the beam source and the object
to be imaged.
2. The method of claim 1, wherein the desired image characteristics
comprise desired levels of contrast-to-noise ratio (CNR),
signal-to-noise ratio (SNR), or a combination thereof.
3. The method of claim 2, wherein the desired image characteristics
providing at least one of: desired image quality in at least one
defined region of interest; and at least one desired distribution
of said image quality.
4. The method of claim 1 wherein performing the calculations
comprises solving an inverse problem using an iterative
solution.
5. The method of claim 1, further comprising defining at least a
first region of interest from a library of population modules or at
least one previously acquired image of the object.
6. The method of claim 1, wherein the total radiation dose to the
patient is lower than that required performing the method in the
absence of intensity modulation or in the absence of the at least a
first collimator.
7. The method of claim 1, further comprising providing at least a
first temporal modulation of the radiation source.
8. The method of claim 1, further comprising providing at least a
first spatial modulation of the radiation source.
9. The method of claim 1, comprising providing both spatial and
temporal modulation of the radiation source.
10. The method of claim 1, wherein the aperture modulator or
collimator comprises a plurality of individual elements adapted to
absorb radiation.
11. The method of claim 10, wherein the plurality of individual
elements are comprised of lead, aluminum, tungsten, a dense
plastic, composite or an alloy, or any combination thereof.
12. An imaging system adapted and configured to perform the method
of claim 1.
13. The imaging system as claimed in claim 12, comprising at least
a first aperture-modulating collimator that comprises a plurality
of individual elements, each being substantially impervious to the
radiation and being movable between an open position and a closed
position, whereby open positions of the individual elements define
an aperture permitting passage of the beam from the radiation
source.
14. The imaging system of claim 12, characterized as a cone-beam
computed tomography system.
15. The imaging system of claim 12, further comprising a
post-processing module for enhancement of at least a first 3-D
model of at least a first region of interest from within the
object.
16. An examination apparatus comprising: an X-ray device for the
generation of X-ray projections of the body volume from different
directions, wherein projections can be based on at least two
different samplings of a collimated beam of X-rays generated from
the device; and the imaging system of claim 12.
17. The examination apparatus according to claim 16, wherein the
X-ray device is adapted and configured to provide an
intensity-modulated, collimated beam of X-rays for imaging at least
a first region of interest of an object examined by the
apparatus.
18. The examination apparatus of claim 16, adapted and configured
as a baggage inspection apparatus, a medical diagnostic apparatus,
a material testing apparatus, or a materials science analytic
apparatus.
19. A record carrier, a computer-readable medium, or a computer
program element wherein: (a) the record carrier comprises a
computer program for the generation of a three-dimensional model of
at least a first region of interest of an object from a plurality
of collimated X-ray projections, and wherein the computer program
is adapted to execute at least one step of a method in accordance
with claim 1; (b) the computer-readable medium comprises a computer
program for reconstructing a three-dimensional image of an object's
region or volume of interest from a set of collimated cone-beam
X-ray projections of the object with an examination apparatus, and
wherein the computer program when being executed by a processor, is
adapted to execute at least one step of a method in accordance with
claim 1; or (c) the computer program element, when being executed
by a processor, is adapted to execute at least one step of a method
in accordance with claim 1.
20. A method for generating a three-dimensional image of a scanned
object from a plurality of cone-beam projections passed through the
object and attenuated thereby, the method comprising: (a)
positioning a source at a position on a predetermined scan path;
(b) passing a projection of cone-beam X-ray radiation comprising a
plurality of projection rays from the source through an object, the
cone-beam projection being attenuated by at least a first
collimator, and by partial absorption in the object; (c) detecting
radiation intensity of the attenuated cone-beam projection on an
area detector; (d) obtaining a two-dimensional attenuation image of
the cone-beam projection from the detected radiation intensity; (e)
generating an intermediate, locally reconstructed,
three-dimensional image with constant values assigned along each
projection ray; (f) at least once repeating steps (d)-(e); and (g)
summing the plurality of intermediate, locally reconstructed,
three-dimensional images obtained for the plurality of cone-beam
projections to obtain an ultimate, reconstructed, three-dimensional
image of the object.
21. A cone-beam tomography apparatus comprising a radiation source,
a radiation detector, a support for an object to be scanned by
radiation from the radiation source, a computer-readable storage
medium storing computer-executable software for generating a
reconstruction of cone-beam radiation attenuation in an object, the
software comprising: code for obtaining and generating an
intermediate, locally reconstructed, three-dimensional image with
constant values assigned along each projection ray; code for
summing the plurality of intermediate, locally reconstructed,
three-dimensional images obtained for the cone-beam projections to
obtain an ultimate, reconstructed, three-dimensional image of the
object; and code for displacing the source and detector relative to
the support in a predetermined scan path for radiation transmitted
from the source, through an object positioned by the support, and
to the detector.
22. A method for forming an image of an object, the method
comprising: (a) exposing the object to a cone beam of radiation
rendered spatially coherent by its passage through at least a first
collimator; (b) projecting the spatially coherent collimated
conical beam of radiation onto the object and collecting the
radiation which has passed through the object in at least a first
detector to produce detected data; (c) deriving, from the detected
data, at least a first image; and (d) repeating step (c) at least
once to form a plurality of images of at least a first region of
interest within the object.
23. A system for forming an image of an object, the system
comprising: (a) a source of a cone beam of radiation rendered
spatially-coherent by at least a first collimator; (b) a detector
for receiving the spatially-coherent radiation which has passed
through the object to produce detected data; and (c) a computer,
receiving the detected data, for deriving from the detected data at
least a first three-dimensional image of at least a first region of
interest within the object.
24. The system of claim 23, characterized as a cone-beam computed
tomography system adapted and configured for medical imaging.
Description
CROSS-REFERENCE TO RELATED APPLICATIONS
[0001] The present application claims priority from U.S.
Provisional Application No. 61/113,655, filed Nov. 12, 2008; the
entire contents of which is specifically incorporated herein by
express reference thereto.
STATEMENT REGARDING FEDERALLY SPONSORED RESEARCH OR DEVELOPMENT
[0002] Not Applicable.
BACKGROUND OF THE INVENTION
[0003] 1. Field of the Invention
[0004] The present invention relates generally to the field of
computed tomography (CT). In particular, the present invention
provides a method of reconstructing a three-dimensional image of an
object's volume of interest from a set of conical-beam,
intensity-modulated projections of this object. Also provided are
image processing devices, examination apparatus, as well as a
computer-readable medium and a program element that is adapted and
configured to perform aspects of the methods disclosed herein.
[0005] 2. Description of Related Art
[0006] Compared to traditional slice-at-a-time tomographic
instruments, the cone-beam (CB) CT offers faster scans, higher
patient throughput, significant reduction in X-ray dosage, and
isotropic resolution. It has a great potential to be applied to a
wide range of medical and industrial applications.
[0007] Cone-beam computerized tomography (CBCT) with flat panel
detectors (FPD) has become a prevalent three-dimensional (3-D)
imaging system for clinical application. Its short time acquisition
and reconstruction with mobile C-arm system allow the integration
of 3-D imaging into the interventional procedures such as
endovascular techniques. In comparison with image intensifiers, the
flat panel detectors provide no geometrical distortion, a higher
dynamic range, and homogeneous signal distribution.sup.1. The
Feldkamp, Davis, Kress (FDK) algorithm.sup.2 and its modifications
in the filtered-backprojection (FBP) family have been implemented
for reconstructing approximate 3-D images from circular CB
projection data. Correction algorithms such as scatter correction,
beam hardening correction, truncation correction, and ring
correction have been developed for removing artifacts of
reconstructed image prior to and after FDK reconstruction
algorithm.sup.3.
[0008] Because an X-ray tube generates X-rays in all directions,
beam restrictors such as collimators or aperture systems are set up
in the X-ray tube for absorbing unnecessary X-ray beams outside
field of view (FOV). Satisfactory CBCT systems employing
collimators to reduce radiation exposure, however, have not been
implemented. As CBCT scans become more common, patients are exposed
to greater amounts of X-ray radiation. This exposure carries an
increased risk for developing cancer, especially for children and
for individuals who require multiple scans. In most cases, the
region of interest for diagnosis is smaller than the window of
exposure to radiation, meaning that healthy surrounding tissue
receives a full dose of radiation despite not being used in the
diagnosis of the patient's condition. Current CT scanning
technology, unlike certain other scanning technologies, does not
allow for shaping the window of exposure because of the complexity
of interpreting less than the full radiation dosage across the
imaging detectors.
BRIEF SUMMARY OF THE INVENTION
[0009] The present invention overcomes these and many other
limitations inherent in the prior art by providing, inter alia,
intensity-modulated conical-beam CT imaging methods that
significantly lower the overall radiation exposure during a
computed tomography (CT) scan. The disclosed methods also provide a
mechanism for interpreting a collimated, intensity-modulated beam,
thus enabling a medical practitioner to narrow the window of
radiation exposure to encompass only the particular region of
interest to be imaged.
[0010] Unlike conventional CT instrumentation that uses constant
beam intensity for every projection angle to produce the final
image, the present invention provides an aperture, such as a
collimator, to modulate the intensity of the X-ray beam, thereby
reducing the amount of radiation impinging on the object of
interest. In the case of medical CT imaging, the methods result is
significantly lower background radiation, and a reduced level of
radiation exposure to surrounding tissues, while still maintaining
a sufficient intensity for imaging the target tissue. CT
instrumentation employing aperture-controlled CB energy provides
overall reduction in radiation exposure, which is expected to
lessen the risk of adverse effects, including carcinogenesis, in
the patient undergoing CBCT imaging and examination.
[0011] In illustrative embodiments, the method is accomplished
using one or more aperture-controlling and/or beam
intensity-restricting or limiting devices, such as, for example, a
plurality of collimators, preferably placed between the radiation
source and the object, to modulate the intensity of the energy
beam. Modulation of the beam intensity using the intensity
modulating device is accomplished such that one or more regions of
interest (ROI) (alternatively referred to as a "volume of interest"
[VOI]) for which analysis and imaging is desired, receive a dose of
radiation sufficient to effect the desired imaging characteristics,
while the surrounding tissue outside the ROI/VOI receives much
lower (i.e., attenuated) ancillary radiation. By narrowing the
target "window" to encompass substantially only the ROI/VOI, the
present invention significantly improves the quality of diagnostic
imaging, while also reducing the risk(s) that are inherent to a
patient undergoing imaging using modalities (such as CT) that
employ ionizing radiation.
[0012] The purposes for applying collimation to CBCT are to reduce
exposing areas of patients that need not be imaged and to moderate
the scatter effect, which degrades image quality.sup.4. Collimators
can be designed as variable diaphragms composed of movable pieces
of metal. They include two movable pieces of metal in the
longitudinal direction (perpendicular to the plane of X-ray source
trajectory) and two in the transverse direction (tangential to the
circle of X-ray source trajectory). The relation between scatter
effect and collimation on different organs of a phantom has been
previously demonstrated.sup.5. The experimental results showed that
scatter-to-primary ratio decreases when FOV in the longitudinal
direction is reduced. Moore et al. calculated the contrast-to-noise
ratio (CNR) of a targeted area using aluminum-made filters and
lead-made collimators with varying thickness and with different fan
angles of FOV in the longitudinal direction.sup.6. The present
invention demonstrates significant improvements in CNR can be
achieved when using collimated beam intensity-modulated imaging
versus traditional non-collimated CT imaging methodologies and
instrumentation.
[0013] In an overall and general sense, CT systems capable of
providing data for reconstructing a three-dimensional (3-D) object
using one or more rotations with an intensity-modulatable beam of
radiation, typically include a source of X-rays distributed along a
line parallel to an axis of rotation, an aperture control
positioned perpendicular to the axis of rotation and near the x-ray
sources so to limit the X-rays illuminating the object to contain
only X-rays that travel substantially along lines perpendicular to
the axis of rotation, and a detector for detecting and measuring
transmitted X-rays emitted by the source. In illustrative
embodiments, the aperture for controlling the size of the CB may
include one or more collimators to restrict the X-rays primarily to
impinge primarily upon the ROI/VOI in the object being imaged.
Exemplary collimating devices may include one or more arrays or
pluralities of collimators, positioned to modulate the intensity of
the radiation source. In certain embodiments, the collimators may
include a plurality of flat collimator plates, blades, lamellas,
gates, vanes, irises, pinholes, or windows, or any combination
thereof, including without limitation, those composed of a piece of
dense plastic, composite, alloy, or metal, including, without
limitation, aluminum, lead, or tungsten, that effectively block at
least a first portion of the beam from reaching the object.
[0014] In a first embodiment, the invention provides a method for
operating computed tomographic imaging using a radiation source and
a plurality of detectors to generate an image of an object. In an
overall and general sense, this method involves one or more steps
including a) defining desired image characteristics; b) performing
calculations to determine the modulation intensity to be applied to
a radiation source by at least a first collimator to generate the
desired image characteristics; and c) modulating the radiation
source using the at least a first aperture system or collimator
device to generate a desired pattern of fluence between the beam
source and the object to be imaged. Exemplary desired image
characteristics include, without limitation, desired levels of
contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), or a
combination thereof. The method may involve providing at least one
desired image quality in at least one defined region of interest,
or alternatively, providing at least a desired plurality or
distribution of images or image qualities. In certain applications,
the step of performing the calculations will include solving an
inverse problem using one or more iterative solutions. Such methods
may also further optionally include a step of defining at least a
first region of interest from a library of population modules, or
from at least one previously acquired image of the object.
[0015] In the practice of these methods upon animal (and
preferably, human) subjects, the total radiation dose to the
organism will be lower (and in some cases substantially lower) than
that typically required performing a similar method in the absence
of intensity modulation, and/or without using at least one
modulatable aperture or collimator device to modulate the intensity
and fluence of the radiation beam emanating from the radiation
source.
[0016] In certain embodiments, the method will include providing at
least a first temporal modulation of the radiation source, at least
a first spatial modulation of the radiation source, or a
combination of both spatial and temporal modulation of the
radiation source.
[0017] In exemplary embodiments, the collimator will include a
plurality of individual elements adapted to absorb radiation, and
in some instances, this plurality of individual elements is
composed of lead or other such like element, compound, or
composition.
[0018] In another embodiment, the invention provides an imaging
system adapted and configured to perform one or more of the methods
described herein. In the practice of the invention, the imaging
system may include at least a first collimator that is composed of
two or more individual elements, each of which is substantially
impervious to the radiation, and being substantially movable
between an open position and a closed position, whereby open
positions of the individual elements within the collimator define
an aperture that permits passage of the beam from the radiation
source to impinge upon the object being imaged. Such imaging
systems include, without limitation, CT systems, and preferably,
CBCT systems, including, without limitation, those CBCT systems in
which a computer forms at least a first image using at least a
first mathematical algorithm adapted and configured for obtaining
and/or analyzing an image of an object of interest.
[0019] In certain embodiments, the imaging system will further
optionally include one or more post-processing modules that may be
useful in the enhancement of at least a first three-dimensional
model of at least a first region of interest from on, or
substantially within one or more portions of the object being
imaged by the system. Preferably, the plurality of X-ray
projections will originate from a circular, or even helical,
trajectory of one or more X-ray radiation sources around, or in
substantial proximity to, the object of interest being imaged.
[0020] In another embodiment, the invention provides a record
carrier on which a computer program for the generation of a
three-dimensional model of at least a first region of interest of
an object from a plurality of collimated X-ray projections is
stored. The record carrier preferably contains a computer program
that is adapted to execute at least one, and preferably,
substantially all of the iterative steps of one or more of the
methods disclosed herein.
[0021] The invention also provides a computer-readable medium, onto
which a computer program for reconstructing one or more
three-dimensional image(s) of an object's ROI/VOI from a set of
collimated CB X-ray projections of the object is stored.
Preferably, the computer program, when being executed by a
processor, is adapted to execute at least one, and substantially
all, of the steps of one or more of the image acquisition and/or
image analysis methods disclosed herein.
[0022] The invention, in another embodiment, also provides a
program element for reconstructing a three-dimensional image of an
object's ROI or VOI from at least a first set of CB X-ray
projections of the object, which, when being executed by a
processor, is adapted to execute at least one step, and
substantially all, of the steps of one or more methods as disclosed
herein.
[0023] In another embodiment, the invention provides an examination
apparatus that includes (a) an X-ray device for the generation of
X-ray projections of an object's region or volume of interest from
at least two different directions, wherein the projections are
obtained from at least two different samplings of a collimated beam
of X-rays generated from the device; and (b) at least a first
imaging acquisition or analysis device as disclosed herein. In one
embodiment, the examination is characterized as a CBCT system, and
is preferably adapted and/or configured to provide at least a first
intensity-modulated (i.e., aperture-limited or collimated) beam of
X-rays for imaging at least a first region of interest of an object
that is undergoing examination in the apparatus by one or more
technicians, operators, medical practitioners, or such like.
[0024] In exemplary embodiments, the examination apparatus as
disclosed herein may be adapted or configured as a baggage
inspection apparatus, a medical diagnostic apparatus, a material
testing apparatus, or a materials science analytic apparatus, or
such like.
[0025] The invention also further herein provides methods for
generating three-dimensional images of an object using a plurality
of CB projections that are passed through the object and attenuated
thereby. In an overall and general sense, such methods include at
least the steps of a) positioning a source at a position on a
predetermined scan path; b) passing a projection of CB X-ray
radiation comprising a plurality of projection rays from the source
through an object, wherein the CB projection is attenuated by at
least one or more collimators adapted and configured to attenuate
the projection; c) detecting the radiation intensity of the CB
projection passing through the object of interest onto an area of
at least a first detector; d) obtaining a two-dimensional
attenuation image of the CB projection from the detected radiation
intensity; e) generating an intermediate, locally reconstructed,
three-dimensional image with constant values assigned along each
projection ray; f) at least once, repeating steps (d)-(e); and g)
summing the plurality of intermediate, locally reconstructed,
three-dimensional images obtained for the plurality of CB
projections to obtain an ultimate, reconstructed, three-dimensional
image of the object.
[0026] The invention also further provides a CBCT apparatus that
includes: (a) at least a first radiation source, (b) at least a
first radiation detector, (c) at least a first support for an
object of interest, or alternatively, at least a first region of
interest from such an object, to be illuminated by a collimated
beam of radiation projected from the radiation source, and (d) at
least one computer-readable storage medium for storing
computer-executable software to generate a reconstruction of the CB
radiation attenuation in the illuminated object. In such
applications, the software will preferably include: (a)
computer-executable code for obtaining and/or generating at least a
first intermediate, locally reconstructed, three-dimensional image;
(b) computer-executable code for summing a plurality of
intermediate, locally reconstructed, three-dimensional images
obtained for the CB projections to obtain an ultimate,
reconstructed, three-dimensional image of the object; and (c)
computer-executable code for displacing the source and detector
relative to the support in a predetermined scan path for radiation
transmitted from the source, through an object positioned by the
support, and to the detector.
[0027] The invention also provides methods and systems for forming
an image of an object that generally involves) exposing the object
to a cone beam of radiation rendered spatially coherent by its
passage through at least a first collimator; (b) projecting the
spatially coherent collimated conical beam of radiation onto the
object and collecting the radiation which has passed through the
object in at least a first detector to produce detected data; (c)
deriving, from the detected data, at least a first image; and (d)
repeating step (c) at least once to form a plurality of images of
at least a first region of interest within the object. Exemplary
systems include, without limitation, CBCT system adapted and
configured for medical diagnostic imaging and/or therapy.
[0028] According to another exemplary embodiment of the present
invention, a computer-readable medium may be provided, in which a
computer program for reconstructing an image from a set of
projections of an object of interest with an examination apparatus
is stored which, when being executed by a processor, is adapted to
carry out the above-mentioned method steps.
[0029] The present invention also relates to a program element of
reconstructing an image from a projection data set of an object of
interest, which, when being executed by a processor, is adapted to
carry out the above-mentioned method steps. The program element may
preferably be loaded into the working memory of a data processor.
The data processor may thus be equipped to carry out exemplary
embodiments of the methods of the present invention. The computer
program may be written in any suitable programming language, and
may be stored on a computer-readable medium, such as a computer
hard drive, a CD-ROM, DVD-ROM, or the like. In addition, the
computer program may be available from a network, from which it may
be downloaded into image processing devices or processors, or any
suitable computer(s).
BRIEF DESCRIPTION OF THE DRAWINGS
[0030] For promoting an understanding of the principles of the
invention, reference will now be made to the embodiments, or
examples, illustrated in the drawings and specific language will be
used to describe the same. It will nevertheless be understood that
no limitation of the scope of the invention is thereby intended.
Any alterations and further modifications in the described
embodiments, and any further applications of the principles of the
invention as described herein are contemplated as would normally
occur to one of ordinary skill in the art to which the invention
relates.
[0031] The following drawings form part of the present
specification and are included to demonstrate certain aspects of
the present invention. The invention may be better understood by
reference to the following description taken in conjunction with
the accompanying drawings, in which like reference numerals
identify like elements, and in which:
[0032] FIG. 1A and FIG. 1B illustrate exemplary ROI imaging with
truncation in accordance with one aspect of the present invention.
In FIG. 1A the ROI imaging has larger FOV, and only several
projections are truncated. The projections undergo mild truncation.
In FIG. 1B, the ROI imaging has smaller FOV, and all projections
are truncated. The projections undergo severe truncation;
[0033] FIG. 2 illustrates an exemplary ROI imaging with movable
collimators in accordance with one aspect of the present
invention;
[0034] FIG. 3A, FIG. 3B, FIG. 3C, and FIG. 3D illustrate an
exemplary truncation correction in accordance with one aspect of
the present invention. In FIG. 3A, the projection image without
truncation and the coordinate definition is shown. In FIG. 3B, the
projection image with truncation, and u.sub.r and u.sub.l are the
edges of the truncated projection are illustrated. In FIG. 3C, the
illustration of extrapolation with mirror symmetry is shown; and in
FIG. 3D the cosine weighting of (c) is shown;
[0035] FIG. 4A and FIG. 4B illustrate an exemplary .pi.-line
reconstruction in accordance with one aspect of the present
invention. In FIG. 4A, the geometry and the trajectory in mid-plane
is shown with z=0. In FIG. 4B, the .pi.-line segments are defined
as the chords on the circular trajectory which lies on the
mid-plane z=0 and imaginary ones (dotted circles) which lies on the
plane z.noteq.0;
[0036] FIG. 5A, FIG. 5B, and FIG. 5C illustrate an exemplary
entrance dose calculation in accordance with one aspect of the
present invention. In FIG. 5A, the virtual entrance plane is
defined at the top of the patients. In FIG. 5B, the geometry of
projection onto the virtual plane, and in FIG. 5C, the obliquity
effect is demonstrated;
[0037] FIG. 6 illustrates an exemplary geometry in the measurement
of radiation dose in accordance with one aspect of the present
invention. The head phantom is placed in the center of the
trajectory with a chamber inserted at the center or at the
peripheral;
[0038] FIG. 7A, FIG. 7B, FIG. 7C, and FIG. 7D show an exemplary FDK
and .pi.-line reconstruction of the mid-plane in accordance with
one aspect of the present invention. In FIG. 7A, the FDK
reconstruction is shown "without collimation" imaging. In FIG. 7B,
the .pi.-line reconstruction is shown "without collimation"
imaging. In FIG. 7C, the FDK reconstruction is shown "with
collimation" imaging with C.sub.t=130 mm and C.sub.l=130 mm. In
FIG. 7D, the .pi.-line reconstruction is shown "with collimation"
imaging C.sub.t=130 mm and C.sub.l=130 mm;
[0039] FIG. 8A, FIG. 8B, and FIG. 8C illustrate an exemplary ROI
reconstruction of the mid-plane in accordance with one aspect of
the present invention. The reconstruction is done by truncation
correction with the FDK algorithm. FIG. 8A shows the results
without collimation. FIG. 8B shows the results with collimation,
C.sub.t=110 mm and C.sub.l=77 mm; FIG. 8C illustrates the results
with collimation, C.sub.t=150 mm and C.sub.l=83 mm;
[0040] FIG. 9 shows an exemplary calculation of entrance dose in
accordance with one aspect of the present invention;
[0041] FIG. 10 shows an exemplary dose comparison of full-field
versus FOV areas in accordance with one aspect of the present
invention;
[0042] FIG. 11A and FIG. 11B illustrate an exemplary relation of
.pi.-line segments and ROI in accordance with one aspect of the
present invention. FIG. 11A illustrates an object under ROI imaging
with two .pi.-line segments, L.sub.1 and L.sub.2. FIG. 11B
illustrates that when the ROI was entirely within the object, every
.pi.-line segment, such as L.sub.3, contains part of the object
outside the ROI; and
[0043] FIG. 12 illustrates a schematic representation of an
exemplary CBCT system in accordance with one aspect of the present
invention.
DESCRIPTION OF ILLUSTRATIVE EMBODIMENTS
[0044] Illustrative embodiments of the invention are described
below. In the interest of clarity, not all features of an actual
implementation are described in this specification. It will of
course be appreciated that in the development of any such actual
embodiment, numerous implementation-specific decisions must be made
to achieve the developers' specific goals, such as compliance with
system-related and business-related constraints, which will vary
from one implementation to another. Moreover, it will be
appreciated that such a development effort might be complex and
time-consuming, but would be a routine undertaking for those of
ordinary skill in the art having the benefit of this
disclosure.
Cone-Beam Computed Tomography (CBCT)
[0045] In CBCT, a set of X-ray cone-beam projections of an object's
ROI is acquired while an X-ray source moves along some source
trajectory around the object. Using a CBCT scanner, a 3-D image of
the object's ROI can be reconstructed from this set of cone-beam
projections. Conventional CBCT scanners are equipped with a
point-like X-ray source and a large-area X-ray detector. Typically,
the detector is flat and subdivided into a two-dimensional (2-D)
array of small detector elements. Such a flat detector naturally
defines a 2-D plane in 3-D space, i.e., the detector plane, which
is rigidly attached to the detector and moves as the detector
moves. The cone-beam is formed by those X-rays that emanate from
the source and intercept the detector. A 3-D image is reconstructed
from the set of cone-beam projections by means of an
image-processing device, which is typically a computer that
executes a computer program that implements a reconstruction
algorithm. The reconstructed image provides a sampled estimate of
the 3-D map of the X-ray attenuation coefficient within the ROI of
the imaged object.
[0046] CBCT scanners may be realized in various ways, which are
known to the person of ordinary skill in the medical imaging arts.
U.S. Pat. Nos. 6,582,120 and 5,124,914 (each of which is
specifically incorporated herein in its entirety by express
reference thereto) illustrate examples of conventional arm-based
and gantry-based CBCT apparatus. C-arm based scanners are normally
equipped with a flat detector, while gantry-based scanners have
either flat or non-flat detectors. Flat detectors have natural
detector planes associated with them that are usually subdivided
into a 2-D array of small detector elements. In such a case, each
CB projection of the acquired set of CB projections will be sampled
on an equidistant Cartesian grid in the detector plane. In non-flat
detector models, a virtual detector plane can be associated with
the detector in which each CB projection can be sampled in the
virtual detector plane. In these detectors, the grid of sampling
points is be planar, but not necessarily equidistant Cartesian.
[0047] CBCT scanners include one or more image processing devices
that reconstruct the acquired images by executing one or more
particular reconstruction algorithms, and typically include one or
more viewing consoles for displaying reconstructed images, and is
configured and adapted to allow an operator, such as a CT
technician or radiologist, to operate the device, as well as obtain
and/or record one or more images obtained by the scanner.
[0048] A number of CT systems, including CBCT, and methods for
collimating, processing and forming an image using such systems are
known in the art, and exemplified, without limitation, in U.S. Pat.
Nos. 7,424,090, 7,418,082, 7,412,029, 7,397,887, 7,396,162,
7,386,088, 7,379,529, 7,375,338, 7,372,937, 7,339,174, 7,317,819,
7,310,411, 7,308,072, 7,305,063, 7,292,717, 7,292,674, 7,289,599,
7,249,886, 7,245,697, 7,233,644, 7,209,535, 7,203,272, 7,187,750,
7,176,466, 7,151,817, 7,147,372, 7,145,981, 7,142,633, 7,116,749,
7,113,570, 7,103,137, 7,101,078, 7,099,428, 7,085,345, 7,076,023,
7,072,444, 7,072,436, 7,058,159, 7,050,534, 7,046,757, 7,042,975,
7,027,558, 7,023,956, 7,020,232, 7,015,477, 7,015,460, 6,865,246,
6,842,502, and 6,735,277; as well as U.S. Patent Appl. Publ. Nos.
20080253516, 20080226018, 20080212860, 20080212735, 20080205727,
20080205587, 20080205585, 20080192886, 20080187195, 20080165923,
20080165918, 20080144904, 20080137802, 20080118032, 20080116386,
20080112531, 20080104530, 20080095305, 20080095301, 20080089468,
20080061241, 20080031507, 20070269001, 20070172026, 20070140410,
20070104310, 20070003010, 20060269049, 20060262898, 20090292898,
20060067464, 20060002506, 20030235265, 20030202637, and
20030043957, the contents of each of which is specifically
incorporated herein in its entirety by express reference
thereto.
Collimation in CBCT
[0049] Collimators in CBCT systems are designed to acquire
customized field of view (FOV) for a smaller object and to reduce
radiation dose and scatter effect. In most applications, only a
region of interest (ROI) within a large body needs to be imaged and
acquired for reducing radiation exposure to patients. However, the
reconstructed ROI image from incomplete projections may contain
severe artifacts.
[0050] In the example that follows, an imaging method is described,
which customizes a small ROI within an object by applying
adjustable collimators and a simple and practical reconstruction by
truncation correction in order to reduce a large amount of
unnecessary radiation dose and to obtain tolerable image quality of
the reconstructed image compared with that without ROI imaging.
[0051] In the studies described herein, ROI imaging was performed
by setting transversely and longitudinally movable lead collimators
between X-ray source and the object. The ROI is constrained to a
small area within an object. Entrance dose has been utilized for
estimating the radiation dose absorbed in the body and conduct
experiments for measuring radiation dose with ROI imaging. Two
reconstruction techniques for possibility of overcoming truncation
problems were investigated: one was the truncation correction
technique with the Feldkamp, Davis, Kress (FDK) algorithm and the
other was .pi.-line reconstruction algorithm.
[0052] The truncation correction technique was applied with FDK to
reconstruct the image of cylindrical polystyrene with water as
background. The imaged regions are small circles with diameters of
70- and 94-mm. From the radiation dose measurement, it was found
that at least 60% and 70% of radiation dose was reduced in each of
ROI imaging. Furthermore, the image quality was still acceptable
with little variation of image gray values; however, it was not
possible using the analytical reconstruction algorithms employed to
fully reconstruct a ROI that was totally encompassed by the
object.
[0053] A truncation correction technique was applied to ROI imaging
with ROI within the object, and a surprising discovery was made: at
least half of radiation dose could be reduced without sacrificing
significant image quality. These results demonstrated that the use
of collimators to reduce radiation dose during CBCT imaging
represents a fundamental change in the way medical CT imaging may
be performed in the future.
Computing Devices and Programmed Media
[0054] The detection, acquisition, analysis, and display of CT
images collected using the disclosed methods and apparatus may
conveniently be implemented using a conventional general purpose
computer or micro-processor programmed according to the teachings
of the present invention, as will be apparent to those skilled in
the computer arts. Appropriate software may readily be prepared by
programmers of ordinary skill based on the teachings of the present
disclosure, as will be apparent to those skilled in the software
arts. Exemplary computers may include a motherboard that contains a
CPU, memory (e.g., DRAM, ROM, EPROM, EEPROM, SRAM, SDRAM,
Flash-RAM, or the like), and other optional special purpose logic
devices (e.g., ASICS, and the like) or configurable logic devices
(e.g., GAL, reprogrammable FPGA, and the like). The computer also
includes plural input devices, including, without limitation, a
keyboard and a mouse, one or more display/graphics card or adaptors
for controlling the output to one or more monitor or display
devices. Additionally, the computer may include removable media
devices, such as, without limitation, a compact disc, magnetic
tape, removable magneto-optical media, and the like); and one or
more fixed, or high-density media drives including, without
limitation, a "hard disk," connected using an appropriate device
bus (e.g., a SCSI bus, an Enhanced-IDE bus, an Ultra-DMA bus, and
the like). The computer may also include a compact disc reader, a
compact disc reader/writer unit, which may be connected to the same
device bus or to another device bus.
[0055] Examples of computer readable media associated with the
present invention include, without limitation, compact discs, hard
disks, magnetic tapes, magneto-optical disks, programmed read-only
memory (e.g., without limitation, EPROM, EEPROM, Flash-EPROM, and
the like), DRAM, SRAM, SDRAM, Rambus-DRAM, etc. and one or more
combinations thereof. Stored on any one or on a combination of
these computer readable media, the present invention includes
software for controlling both the hardware of the computer and for
enabling the computer to interact with a human user. Such software
may include, without limitation, one or more device drivers, one or
more operating systems and one or more user applications, including
development tools and the like. Computer program products of the
present invention include any computer readable medium which stores
computer program instructions (e.g., computer code devices) which
when executed by a computer causes the computer to perform the
method of the present invention. The acquired data may be
digitized, if not already in digital form, and alternatively, the
source of image data being obtained and processed may be a memory
storing data produced by an image acquisition device, and the
memory may be local or remote, in which case a data communication
network may be used to access the image data for processing
according to the present invention.
EXAMPLE
[0056] The following example is included to demonstrate
illustrative embodiments of the invention. It should be appreciated
by those of skill in the art that the techniques disclosed in the
example that follows, represent techniques discovered by the
inventors to function well in the practice of the invention, and
thus can be considered to constitute preferred modes for its
practice. However, those of ordinary skill in the art should, in
light of the present disclosure, appreciate that many changes can
be made in the specific embodiments which are disclosed and still
obtain a like or similar result without departing from the spirit
and scope of the invention.
Example 1
Region of Interest Reconstruction and Dose Reduction Estimation in
Collimated CBCT Imaging
[0057] When objects or bodies extend outside scan FOV, there is
abrupt discontinuity at the edges of detector, and projection data
is truncated since missing data outside the detector is
theoretically assumed to be zero during CBCT reconstruction. In
practice, nearly all of the reconstruction algorithms in CBCT
scanners are based on the FDK algorithm. Performing one-dimensional
(1-D) filtering on truncated projection data leads to bright band
artifacts extending inside scan FOV and incorrect reconstruction
around the edges of scan FOV.sup.7. FIG. 1A shows an example of
region of interest (ROI) imaging with mild truncation on projection
data commonly encountered due to limitation of detector size. The
object is entirely visible at several views, but truncation occurs
at the other views. Several analytical reconstruction
algorithms.sup.8,9 in the backprojection-filtration (BPF) family
can exactly reconstruct ROI images from projections with mild
truncation. As ROI is completely included in the object (FIG. 1B)
and undergoes severe truncation, projection images at all views
contain truncations. This case is well known as an interior problem
and is not solvable by analytical approaches. An alternative
possible method is the iteration reconstruction, in which forward
projection is formulated as
Wf=p, (1)
[0058] where f represents the image vector, p the projection
vector, and W the weighting matrix. Equation (1) is expressed as a
linear equation from line integral (Radon transform). The image
vector f can be obtained by solving the inverse problem in least
square sense. As described previously, biased but informative
solutions can be iteratively computed.sup.10. However, it lacks
computation speed, and due to severe truncation, little information
in p is available to increase accuracy of the reconstructed image,
f.
[0059] A number of truncation correction techniques commonly
implemented on mildly truncated projections as the case in FIG. 1A
have been proposed using extrapolation on missing data. Ohnesorge
et al. has proposed a symmetric mirroring extrapolation at the
edges of truncated projections.sup.11. After extrapolation, a
cos-type weighting which smoothens transition of extrapolated data
to the base line was applied. Hsieh et al. extrapolated the
truncated projections, because total attenuation for each
projection would theoretically remain constant as if there was no
truncation.sup.12. The scanned object was assumed to be a water
cylinder, and the magnitudes and slopes at the edge of truncated
projections were used to estimate the portion of the water cylinder
that extends beyond the original projections within FOV and fits
missing parts. The amount of extended projection was adjusted
according to the constant total attenuation estimated at
non-truncated scan views. Other models for extrapolation such as a
square root of quadratic functions.sup.13 and linear
prediction.sup.14,15 were explored.
[0060] When changes of 3-D images are confined to a small region
during interventions, ROI imaging is applied during the second scan
in order to save radiation dose. Projections are generally
truncated at all views as in FIG. 1B. Missing data during the
second scan of ROI imaging can be approximately recovered from the
previously acquired projections .sup.16. The preliminary
reconstructed ROI is registered to the previously acquired 3-D
images. Similar studies using a priori information for
reconstruction with limited FOV were previously discussed.sup.17
18.
[0061] In this example, ROI imaging technique uses movable pieces
of lead as collimators. The collimators make the scan FOV reduced
such that the imaged region is as small as a circle having a
diameter of approximately 70-mm, and all projections were under
severe truncation at all views as shown in FIG. 1B. Dose
measurement studies were conducted to demonstrate that at least
half of radiation dose was reduced. The 3-D reconstruction was
performed by truncation correction using the FDK algorithm as noted
above. The object in ROI could still be reconstructed with
acceptable image quality, and the data demonstrated that the
disclosed method was a straightforward way of dealing with the
challenging interior problem from a practical perspective.
Furthermore, an analytical reconstruction method, called a
.pi.-line reconstruction in the BPF family, has been examined under
severe truncation situation. The simulation revealed that the
.pi.-line reconstruction algorithm without truncation correction
still produced the same bright artifacts as FDK does.
Materials and Methods
[0062] ROI Imaging with Collimators
[0063] FIG. 2 illustrates the geometry of an exemplary CBCT system
using collimators, and shows how such collimators may be employed
in CBCT imaging. The collimators were set up between the object and
the source, and could be adjusted transversely and/or
longitudinally to customize an open rectangular area according to a
particular ROI size. A lead shield having a thickness of about 1.1
mm can attenuate 100-kV X-ray beams to 0.1%.sup.19. If the
collimators are made of lead and thick enough, X-ray beams outside
the open area are greatly attenuated and can be ignored. Hence, no
information is available on projections outside FOV for
reconstruction with lead collimation.
[0064] ROI Imaging Truncation Correction
[0065] First, the truncation correction technique is as designed as
follows: letting P.sub.k(u,v) be the k-th projection image with u-v
coordinate shown in FIG. 3A, it was assumed that it has been
converted to accumulated attenuation projection from
intensity-based one on the detector. Assume the reconstruction
algorithm is based on FBP with 1-D filtering in u-coordinate, which
is defined as the transverse direction. For simplicity, the index k
of P(u,v) was ignored. The projection truncated due to reduced FOV
(as illustrated in FIG. 3B) is denoted as:
P FOV ( u , v ) = { 0 u > u r 0 u < u l P ( u , v ) u l
.ltoreq. u .ltoreq. u r , ( 2 ) ##EQU00001##
[0066] where u.sub.r and u.sub.l are determined by the right and
left edges of the truncated projection P.sub.FOV(u,v). Assuming the
lengths for extrapolations at the both sides are N.sub.r and
N.sub.l with 0<N.sub.l,N.sub.r<u.sub.r-u.sub.l, based on the
symmetric mirroring extrapolation technique.sup.11, the
extrapolated projection (FIG. 3C) is:
P ex ( u , v ) = { ( 2 P FOV ( u r , v ) - P FOV ( - u + 2 u r , v
) ) , N r + u r .gtoreq. u > u r ( 2 P FOV ( u l , v ) - P FOV (
- u + 2 u l , v ) ) , u l - N l .ltoreq. u < u l P FOV ( u , v )
, u l .ltoreq. u .ltoreq. u r 0 , otherwise . ( 3 )
##EQU00002##
[0067] By assuming that the image variation in ROI is small
compared with the magnitudes P.sub.FOV(u.sub.r,v) and
P.sub.FOV(u.sub.l,v), the assumption, in general, is held for small
ROI imaging of a body or head. Therefore, the extrapolated
projection P.sub.ex(u,v) is positive under this assumption:
P.sub.FOV(u.sub.r,v)>|P.sub.FOV(u.sub.r,v)-P.sub.FOV(-u+2u.sub.r,v)|
P.sub.FOV(u.sub.l,v)>|P.sub.FOV(u.sub.l,v)-P.sub.FOV(-u+2u.sub.l,v)|
[0068] To obtain smooth transition of the extrapolated data to
zero, P.sub.ex(u,v) is weighted by Equation (4):
w ( u , v ) = { 1 2 ( 1 + cos ( .pi. ( u - u r ) N r ) ) u > u r
1 u l .ltoreq. u .ltoreq. u r 1 2 ( 1 + cos ( .pi. ( u - u l ) N l
) ) u < u l . ( 4 ) ##EQU00003##
[0069] The weighted extrapolation is illustrated in FIG. 3D.
[0070] .pi.-Line Reconstruction Algorithm
[0071] The .pi.-line reconstruction algorithm is described below:
FIG. 4A shows the geometry and the trajectory of circular CBCT in
the mid-plane. .theta. was defined as the gantry rotational angle
of the source starting from y-axis counterclockwise. The concept of
a .pi.-line originated from helical CBCT reconstruction
algorithms.sup.20,21, but the modification in circular CBCT
geometry was first described by others.sup.22. FIG. 4B shows the
3-D view of the gantry rotation and .pi.-line segments. The
trajectory of the source lies on z=0 plane, and the imaginary
trajectories (dotted circles) on the planes with z.noteq.0. The
.pi.-line segments are defined as parallel chords on these circular
trajectory and imaginary ones. Letting r=(x.sub.r,y.sub.r,z.sub.r)
be a point on the .pi.-line segment CD of the z=z.sub.r plane and
r(.theta..sub.A) and r(.theta..sub.B) points A and B on the
trajectory, where .theta..sub.A and .theta..sub.B are the
rotational angles of the source at A and B, all parallel .pi.-line
segments have the same direction defined by a unit vector:
u .pi. = r ( .theta. A ) - r ( .theta. B ) r ( .theta. A ) - r (
.theta. B ) . ( 5 ) ##EQU00004##
[0072] Assuming that the .pi.-line segment AB has the same x and y
coordinates as the CD except that they are at different z planes,
the point r in 3-D space with a specific .pi.-line direction
u.sub..pi. can be uniquely determined by a quadruple
(.lamda.,.theta..sub.A,.theta..sub.B,z.sub.r) as
r=r(.theta..sub.A)+z.sub.ru.sub.z+.lamda.u.sub..pi. (6)
[0073] where .lamda..epsilon.[0,1], and u.sub.z=(0,0,1). The
reconstruction has two steps: first, back-project the projections
into the .pi.-line segments, and secondly, perform Hilbert
transformation along each of .pi.-line segments. The 3-D volume is
union of .pi.-line segments and, for simplicity, we describe the
practical reconstruction of 3-D images only on a .pi.-line segment
based on theoretical results previously described.sup.21.
[0074] Considering the .pi.-line segment, CD, and letting
P.sub..theta.(u,v) be the projection image at some rotational angle
.theta. with the same u-v coordinate mentioned above, and letting L
be the distance between the source and the detector, by Equation
(6), r is equivalent to
(.lamda.,.theta..sub.A,.theta..sub.B,z.sub.r). The back-projected
image b(.lamda.,.theta..sub.A,.theta..sub.B,z.sub.r) or b(r) on the
.pi.-line segment CD is expressed as:
b ( .lamda. , .theta. A , .theta. B , z r ) = .intg. .theta. A
.theta. B L 2 ( R + x sin .theta. - y sin .theta. ) 2
.differential. .differential. u ( R u 2 + v 2 + L 2 P .theta. ( u ,
v ) ) ( u , v ) = ( u r , v r ) .theta. + P .theta. B ( u r , v r )
d ( r , .theta. B ) - P .theta. A ( u r , v r ) d ( r , .theta. A )
. ( 7 ) ##EQU00005##
where d(r,.theta.) was defined as the distance between r and
r(.theta.). P.sub..theta.(u.sub.r,v.sub.r) is the line-integral
which starts from the source at the rotational angle .theta.,
passes through the point r, and ends at a point (u.sub.r,v.sub.r)
on the detector.
[0075] Finally, the reconstructed image along the .pi.-line segment
CD is:
f ^ ( .lamda. , .theta. A , .theta. B , z r ) = 1 2 .pi. H { (
.lamda. B - .lamda. ) ( .lamda. - .lamda. A ) b ( .lamda. , .theta.
A , .theta. B , z r ) } + 1 2 .pi. ( .lamda. B - .lamda. ) (
.lamda. - .lamda. A ) ( P .theta. A ( u c , v c ) + P .theta. B ( u
c , v c ) ) , ( 8 ) ##EQU00006##
where c is the midpoint of CD, and H{a(.lamda.)} is the Hilbert
transform of a(.lamda.). [.lamda..sub.A,.lamda..sub.B] is an
interval such that
{r(.theta..sub.A)+z.sub.ru.sub.z+.lamda.u.sub..pi.|.lamda..epsilon.[.lamd-
a..sub.A,.lamda..sub.B]} includes support set of the object and
belongs to the segment CD.
[0076] Calculation of Radiation Dose Reduction
[0077] An "entrance dose" of X-rays is the dose absorbed at surface
of the skin where X-ray beams enter. An X-ray beam enters the body
from the direction of the X-ray tube. A small share of the beam
exits from the body on the opposite side, where it exposes the
detector. The share of the beam that never exits from the body is
absorbed as extra energy by the body's internal organs and bones.
Between the collimators and the patient, imagine that there is a
virtual plane parallel to the detector and close to the top of the
patient illustrated in FIG. 5A. It could be viewed as a virtual
entrance to the patient X-ray beams are about to enter. The
entrance dose is thus defined as energy of X-ray beams passing
through the virtual plane.
[0078] Letting X-ray intensity be a unit of energy per unit area,
and assuming that the energy which is obtained by integrating the
X-ray beam intensity with a small sphere surrounding the X-ray
source is E.sub.s in FIG. 5B, the source-to-virtual-plane distance
be d, O the origin of the coordinate system of the virtual plane,
and r the distance between the source and a arbitrary point Q on
the virtual plane. Thus, the intensity I.sub.O at the point O is
given by
I O = E s 4 .pi. d 2 , ( 9 ) ##EQU00007##
where 4.pi.d.sup.2 is the surface area of a sphere with radius d.
Assuming that the coordinate of the point Q is (x,y), then the
intensity I.sub.Q at the point Q is
I Q = E s 4 .pi. ( d 2 + x 2 + y 2 ) . ( 10 ) ##EQU00008##
[0079] From Equation (9) and Equation (10), I.sub.Q becomes
I Q = I O d 2 ( d 2 + x 2 + y 2 ) . ( 11 ) ##EQU00009##
[0080] The obliquity effect occurs since surface of the detector is
not perpendicular to the direction of X-ray beams propagation. In
FIG. 5C, the X-ray beams vertically pass through the origin O of
the virtual plane, and denote the cross-section unit area of the
X-ray beams as A.sub.O. The cross-section area of the X-ray beams
on the virtual plane becomes A.sub.Q=A.sub.O/cos .theta. when the
same X-ray beams with cross-section unit area A.sub.O are projected
onto the point Q, with an angle, .theta.. The measured intensity
I.sub.Q due to obliquity alone is
I.sub.Q'=I.sub.Q cos .theta. (12)
where cos
.theta. = d r . ##EQU00010##
[0081] The combination of inverse square law and obliquity effect
is multiplicative. From Equation (11) and Equation (12), the
overall intensity at the point Q is given by:
I Q ' = I O d 3 ( d 2 + x 2 + y 2 ) 3 2 . ( 13 ) ##EQU00011##
[0082] Letting the area projected on the virtual plane within FOV
be A.sub.FOV shown in FIG. 5A, the entrance dose D.sub.E was
calculated by integrating I.sub.Q' over the area A.sub.FOV as set
forth in Equation (14):
D E = .intg. .intg. A FOV I Q ' A . ( 14 ) ##EQU00012##
[0083] Letting A.sub.FOV,m be the maximal area on the virtual plane
when full-field imaging is applied, and denoting the entrance dose
as D.sub.E,m, entrance dose was defined relative to that in the
full field imaging in Equation (15) as:
D % = D E D E , m .times. 100 % , ( 15 ) ##EQU00013##
i.e., D.sub.% is the entrance dose relative to the maximal entrance
dose D.sub.E,m.
[0084] Measurement of Radiation Dose
[0085] Dose measurement in the study was made using a cylindrical
CT dose-index (CTDI) phantom with a diameter of 160 mm. This is the
standard Perspex.RTM. CTDI head phantom with a central hole and
four peripheral holes 10-mm below the surface. FIG. 6 shows the
head phantom during CBCT scans from the longitudinal view and the
two locations at which a CT chamber is inserted (two solid points).
The CT chamber was inserted either in the center or in the
peripheral position during each CBCT scan. The gantry rotation was
over .beta.=-90.degree..about.130.degree.. The measured doses, R,
are given in Roentgen units. The relationship between measured
doses R and CTDI.sub.100 is given by Equation (16):
CTDI.sub.100=0.876.times.2.times.R (16)
[0086] where 2 is a modified factor for the specific CT chamber in
the measurement, and 0.876 is a factor converting Roentgen units
into rads in air. The weighted dose CTDI.sub.w was also calculated
from Equation (17) as:
CTDI.sub.w=1/3CTDI.sub.100,c+2/3CTDI.sub.100,p, (17)
where CTDI.sub.100,c and CTDI.sub.100,p are measured doses at the
center and at the peripheral, respectively.
Results
[0087] The CBCT imaging of the described studies were based on the
C-arm flat panel (FP) CT system (Siemens Medical Solutions,
Forchheim, Germany), and the parameters of CBCT system are given in
Table 1. The collimators were made of lead with a thickness of 5
mm, which was enough to attenuate the intensity of X-ray beams
close to zero. Two collimator pieces are movable in the transverse
direction, while the two remaining collimator pieces are movable in
the longitudinal direction as illustrated in FIG. 2. When the open
area is large enough to allow the X-ray beams to be projected onto
the full field of the detector, the full field imaging was denoted
as being "without collimation." When the collimators are adjusted
to customize a smaller open area so that only ROI imaging is
applied, this was denoted as being "with collimation." In the "with
collimation" imaging, the transverse length and longitudinal width
of rectangular projection images are defined as C.sub.t and
C.sub.l, respectively.
[0088] FDK & .pi.-Line Reconstruction Under Severe Truncation
Correction
[0089] Using a head phantom with tissue- and bone-like materials as
an object, two studies were conducted: the first was "without
collimation" imaging, and the other was "with collimation" imaging
with C.sub.t=130 mm and C.sub.l=130 mm. The reconstructions were
the FDK algorithm and .pi.-line BPF algorithm, each of which is
applied in both studies. The projection images were stored as
integers within the range [32,4096], and were scaled to [0,1]. FIG.
7 shows the reconstructed images with a display window
[0.002,0.007] in the mid-plane (z=0). In FIG. 7B and FIG. 7D, only
the ROI within FOV are displayed, and the bright artifacts are
produced under severe truncation.
[0090] ROI Reconstruction
[0091] Cylindrical polystyrene (about -30 HU) was used with a
diameter of 30 mm and height of 25 mm as the object in the ROI and
water as the background. The cylindrical polystyrene was stacked
above a cylindrical plastic water phantom (about 30 HU) in a tank
of water. These materials and background provide constant
attenuation coefficient that clearly demonstrated the variation of
ROI images in the "with collimation" imaging. The projection of the
cylindrical polystyrene is approximately in the isocenter of the
detector at all views.
[0092] Three studies were performed: one "without collimation"
imaging, another "with collimation" imaging with C.sub.t=150 mm and
C.sub.l=83 mm and the third "with collimation" imaging where
C.sub.t=110 mm and C.sub.l=77 mm. The later study "with
collimation" imaging was conducted with the smallest FOV, which
enabled the projection of the cylindrical polystyrene to exactly
fit within FOV at all views. The CBCT 3-D reconstruction used the
FDK algorithm with ring correction, scatter correction, and
overexposure correction in the CBCT system. In the "with
collimation" imaging, the truncation correction described above was
applied on projection images before FDK reconstruction. The
extrapolation lengths N.sub.r and N.sub.l were assumed to be
one-half C.sub.t. FIG. 8 shows the reconstruction of these three
studies with Hounsfield units on the mid-plane (z=0), on which the
circular trajectory of CT scan lies. Only the ROI imaged within FOV
in FIG. 8B and FIG. 8C is displayed.
[0093] Simulation for Entrance Dose
[0094] In these simulations, the calculation of entrance dose was
based on parameters of the C-arm flat panel (FP) CT system shown in
Table 1. The estimated source-to-virtual-plane distance was 650 mm,
and the estimated A.sub.FOV,m 35213.75 mm.sup.2 (162.5.times.216.7
mm.sup.2), which was A.sub.FOV when the "without collimation"
imaging is applied. FIG. 9 shows the D.sub.% versus A.sub.FOV with
a fixed ratio of width-to-length of 3:4, and a logarithmic (base
10) scale was used for the A.sub.FOV axis.
[0095] Radiation Dose Measurement with Collimation
[0096] The radiation doses were measured with varying open areas of
transverse and longitudinal collimators. The size
C.sub.t.times.C.sub.l of full field ("without collimation") was
385.times.295 mm.sup.2, and its measured doses, tube voltages, and
tube current at the center and at the peripheral were 2.35
Roentgen, 79 kV, and 251 mA; and 2.03 Roentgen, 79 kV, and 251 mA,
respectively. Two studies were conducted: one only varied the
length C.sub.t with the fixed C.sub.l=295 mm, while the other
varied both length C.sub.t and width C.sub.l. The doses, tube
voltages, and tube currents at the center and peripheral were
measured in these two studies as shown in Table 2 and Table 3.
[0097] It is worth noticing that the doses were not lower at the
smallest open areas (C.sub.l=25 mm) of collimators in Table 2 and
Table 3. The tube voltages and currents for C.sub.l=25 mm in Table
2 and Table 3 were above 90 kV and below 220 mA compared with the
others at about 80 kV and 250 mA. This can be explained by the
automatic exposure control in the CBCT system. The tube voltage of
X-ray beams would be automatically increased for fewer photons
detected on the detector. Due to the exposure control mechanism, it
was not possible to use the same tube voltage and current for all
the measurements. It was also not possible to scale the doses to
those with fixed tube voltage and current due to lack of
relationship of dose and tube parameters under collimated imaging.
For the other measurements, the variation of tube voltages and
currents was within 5% relative to 80 kV and 250 mA, and the
measured doses were still reliable in this study.
[0098] The dose for full-field (i.e., "without collimation") was
denoted as 100%. FIG. 10 shows the dose relative to full-field
(i.e., "without collimation") versus varying areas within FOV, and
illustrate the studies as performed according to Table 2 and Table
3. A logarithmic (base 10) scale was used for the axis of areas. As
mentioned above, the doses for C.sub.l=25 mm in FIG. 10 were lower
than those where C.sub.l=85 mm.
[0099] The main purpose of simulating the two analytical
reconstruction algorithms (FDK and .pi.-line) in FIG. 7 was to
explain the impossibility of ROI reconstruction under the severe
truncation defined in FIG. 1B. FIG. 11A illustrates an object under
ROI imaging with two .pi.-line segments, L.sub.1 and L.sub.2. It
could be observed that L.sub.1 passes through the support set of
the object in ROI, but part of intersection of L.sub.2 with the
object was not within the ROI. During the .pi.-line reconstruction,
each projection contained sufficient information for
back-projection onto the L.sub.1 regardless of partial truncation,
but that was not the case for L.sub.2. It was not possible to
obtain correct reconstruction along L.sub.2 when performing
filtering on incompletely back-projected images, which explains why
artifacts appeared only when there was missing data in the filtered
images. When the ROI was entirely within the object (as in FIG.
11B), every .pi.-line segment, such as L.sub.3, contains part of
the object outside the ROI. Like FBP analytical reconstruction
algorithms, it is impossible for BPF reconstruction algorithms to
avoid filtering on insufficient data when the ROI is inside the
object.
[0100] In the studies involving ROI reconstruction, the object of
interest is the cylindrical polystyrene with water as the
background. Two studies were performed using moderate (FIG. 8B) and
smallest (FIG. 8C) FOV to demonstrate feasibility of a proper 3-D
reconstruction even using severe truncation. The gray values
approaching the border of ROI in the "with collimation" imaging
were distributed less uniformly than the "without collimation"
imaging, but the variation of gray values was acceptable for
recognition of the ROI image. From FIG. 10, it was determined that
the weighted doses were decreased by at least 60% and 70% in both
"with collimation" studies. This substantial reduction in radiation
dose enables one now to rethink collimated ROI imaging with only
slight, or insignificant, loss of image quality. Exploiting the
limit of ROI, an object the size of a coin was small enough to
demonstrate the possibility of ROI reconstruction with even the
smallest fields of view.
[0101] It was observed that the extrapolation technique in
truncation correction with the FDK reconstruction could largely
recover ROI image with an extremely small ROI. The various filter
kernels in the FDK algorithms were derived from the ramp filter,
which is formed by Sinc function. The feature of the filters is
that their impulse responses decay fast toward both sides. Since
the FDK reconstruction is involved with convolution of the filter
and projections, the reconstructed images toward isocenter do not
undergo significant distortion. If one properly extends the
truncated projections by extrapolation, it mitigates the variation
of reconstructed images caused by filtering truncated projections.
The distortion in gray values becomes more obvious for
reconstructed images far away from isocenter. Therefore, if a
larger angle of FOV is applied, the distortion of the object in ROI
will be lessened. This effect was clearly demonstrated in FIG. 8B
and FIG. 8C, although it was observed that there still exists a
trade-off between image quality and the angle of FOV.
[0102] The radiation dose calculation in computer simulations can
be performed by Monte Carlo calculations in which photoelectric
effect and scattering of photons are randomly chosen according to
the probability distribution of the mass attenuation coefficient of
a voxel in a simulated object. Validation of the Monte Carlo
simulation has been described in detail elsewhere.sup.23-25.
However, the dose calculation is still impractical, since the
calculated dose distribution varies from object to object using the
same CT system. The calculation of entrance dose in FIG. 9 provides
a good representation for estimating the reduction in radiation
dose in the collimated CT imaging. The amount of entrance dose may
serve as an appropriate indicator of how much radiation dose the
patient would be potentially exposed to.
TABLE-US-00001 TABLE 1 PARAMETERS IN EXEMPLARY CBCT SYSTEMS
Source-to-detector distance 1200 mm Source-to-isocenter distance
750 mm Size of the detector 300 .times. 400 mm.sup.2 Size and
resolution of projection images 1240 .times. 960 (0.308 mm .times.
0.308 mm) Number of projections 543 Scan angle 220.degree. Scan
time 20 s
TABLE-US-00002 TABLE 2 MEASURED DOSES, TUBE VOLTAGES, AND TUBE
CURRENTS WITH FIXED C.sub.l = 295 mm Center Peripheral C.sub.t Dose
kV MA Dose kV mA 25 mm 1.82 92 216 0.74 92 217 85 mm 1.75 78 257
1.10 78 256 145 mm 2.18 80 249 1.71 80 249 205 mm 2.21 80 251 1.85
80 250 265 mm 2.23 80 250 1.95 79 251 325 mm 2.23 80 250 1.96 79
251
TABLE-US-00003 TABLE 3 MEASURED DOSES, TUBE VOLTAGES, AND TUBE
CURRENTS WITH VARYING SIZES Center Peripheral C.sub.t .times.
C.sub.l Dose kV mA Dose kV mA 25 .times. 25 mm.sup.2 0.743 98 203
0.343 97 204 85 .times. 85 mm.sup.2 0.422 80 250 0.571 83 240 145
.times. 145 mm.sup.2 0.861 77 260 0.72 81 247 205 .times. 205
mm.sup.2 1.49 80 248 1.35 80 249 265 .times. 265 mm.sup.2 1.70 80
250 1.63 76 261 325 .times. 295 mm.sup.2 1.89 79 251 1.76 79
251
REFERENCES
[0103] The following references, to the extent that they provide
exemplary procedural or other details supplementary to those set
forth herein, are specifically incorporated herein in their
entirety by explicit reference thereto: [0104] 1. Yu R., Conover D.
"Flat panel detector-based cone-beam volume CT angiography imaging:
system evaluation," Medical Imaging, IEEE Transactions, 19:949-963,
2000. [0105] 2. Feldkamp L. A., Davis L. C., Kress J. W. "Practical
cone-beam algorithm," J. Opt. Soc. Am., 1:612-619, 1984. [0106] 3.
Zellerhoff M., Scholz B., Ruehrnschopf E. P., Brunner T. "Low
contrast 3D reconstruction from C-arm data." Proc. SPIE, 5745:646,
2005. [0107] 4. Prince J. L., Links J. M. "Medical imaging signals
and systems," Upper Saddle River, NJ: Pearson Prentice Hall, 2006.
[0108] 5. Siewerdsen J. H., Jaffray D. A. "Cone-beam computed
tomography with a flat-panel imager: Magnitude and effects of X-ray
scatter," Med. Phys., 28:220, 2001. [0109] 6. Moore C. J., Marchant
T. E., Amer A. M. "Cone beam CT with zonal filters for simultaneous
dose reduction, improved target contrast and automated set-up in
radiotherapy," Phys. Med. Biol., 51:2191-204, 2006. [0110] 7. Gore
J. C., Leeman S. "The reconstruction of objects from incomplete
projections," Phys. Med. Biol., 25:129-136, 1980. [0111] 8. Yu L.,
Zou Y., Sidky E. Y., Pelizzari C. A., Munro P., Pan X. "Region of
interest reconstruction from truncated data in circular cone-beam
CT," Medical Imaging, IEEE Transactions 25:869-881, 2006. [0112] 9.
Noo F., Clackdoyle R., Pack J. D. "A two-step Hilbert transform
method for 2D image reconstruction," Phys. Med. Biol.,
49:3903-3923, 2004. [0113] 10. Zhang B., Zeng G. L.
"Two-dimensional iterative region-of-interest (ROI) reconstruction
from truncated projection data," Med. Phys., 34:935, 2007. [0114]
11. Ohnesorge B., Flohr T., Schwarz K., Heiken J. P., Bae K. T.
Efficient correction for CT image artifacts caused by objects
extending outside the scan field of view. Med. Phys., 27:39, 2000.
[0115] 12. Hsieh J., Chao E., Thibault J. et al. "A novel
reconstruction algorithm to extend the CT scan field-of-view," Med.
Phys., 31:2385, 2004. [0116] 13. Sourbelle K., Kachelriess M.,
Kalender W. A. Reconstruction from truncated projections in CT
using adaptive detruncation. Eur. Radiol., 15:1008-1014, 2005.
[0117] 14. Rajgopal K., SrlNlVasa N., Ramakrishnan K. R. Image
reconstruction from incomplete projection data: A linear prediction
approach. Medical Imaging Systems Techniques and Applications,
1997. [0118] 15. Anoop K. P., Rajgopal K. Estimation of Missing
Data using Windowed Linear Prediction in Laterally Truncated
Projections in Cone-Beam CT. Engineering in Medicine and Biology
Society, 2007. EMBS 2007. 29th Annual International Conference of
the IEEE 2903-2906, 2007. [0119] 16. Wiegert J., Bertram M., Netsch
T., Wulff J., Weese J., Rose G. "Projection Extension for Region of
Interest Imaging in Cone-Beam CTI," Acad. Radiol., 12:1010-1023,
2005. [0120] 17. Ramamurthi K., Prince J. "Tomographic
Reconstruction for Truncated Cone Beam Data Using Prior CT
Information. Medical Image Computing and Computer-Assisted
Intervention"-Miccai 2003: 6th International Conference, Montreal,
Canada, November 2003: Proceedings 2003. [0121] 18. Ruchala K. J.,
Olivera G. H., Kapatoes J. M., Reckwerdt P. J., Mackie T. R.
"Methods for improving limited field-of-view radiotherapy
reconstructions using imperfect a priori images," Med. Phys.,
29:2590, 2002. [0122] 19. Hubbell J. H. "Photon mass attenuation
and energy-absorption coefficients from 1 keV to 20 MeV," Int. J.
Appl. Radiat. Isotop., 33:1269-1290, 1982. [0123] 20. Danielsson P.
E., Edholm P., Seger M. "Towards exact 3D-reconstruction for
helical cone-beam scanning of long objects. A new detector
arrangement and a new completeness condition," Proc. 1997 Meeting
on Fully 3D Image Reconstruction in Radiology and Nuclear Medicine
(Pittsburgh)(D W Townsend and P E Kinahan, eds.) 141-44, 1997.
[0124] 21. Zou Y., Pan X. "Exact image reconstruction on PI-lines
from minimum data in helical cone-beam CT," Phys. Med. Biol.,
49:941-959, 2004. [0125] 22. Pan X., Xia D., Zou Y., Yu L. "A
unified analysis of FBP-based algorithms in helical cone-beam and
circular cone- and fan-beam scans," Phys. Med. Biol., 49:4349-4369,
2004. [0126] 23. Deak P., van Straten M., Shrimpton P. C., Zankl
M., Kalender W. A. "Validation of a Monte Carlo tool for
patient-specific dose simulations in multi-slice computed
tomography," Eur. Radiol., 18:759-772, 2008. [0127] 24. Schmidt B.,
Kalender W. A. "A fast voxel-based Monte Carlo method for scanner-
and patient-specific dose calculations in computed tomography,"
Physica. Medica., 18:43-53, 2002. [0128] 25. Sechopoulos I.,
Vedantham S., Suryanarayanan S., D'Orsi C. J., Karellas A. "Monte
Carlo and Phantom Study of the Radiation Dose to the Body from
Dedicated CT of the Breast," Radiology, 247:98, 2008.
[0129] All of the methods disclosed and claimed herein can be
executed without undue experimentation in light of the present
disclosure. While the methods of this invention and devices,
imaging systems, and computer program products employing them have
been described in terms of exemplary embodiments, it will be
apparent to those of ordinary skill in the art that variations may
be applied to the composition, methods and in the steps or in the
sequence of steps of the method described herein without departing
from the concept, spirit and scope of the invention as defined by
the appended claims. Accordingly, the exclusive rights sought to be
patented are as set forth in the following claims:
* * * * *